基于有色Petri网的750kV变电站故障诊断研究

发布时间：2018-07-13 11:38

【摘要】：西北750kV联网工程是一项超高压输变电工程,它具有规模大、线路长、电压高、机组大、联网程度紧密等特点,是西北地区输送各种形式电能关键部分。750kV变电站作为其核心纽带作用,智能化程度越来越高,监控装置越来越多,其采用双网双保护配置,在拓扑结构和保护配置相对于一般的变电站和输电网都复杂和繁琐。一旦发生故障,会涌入大量的报警信息,其信息的种类繁多、冗余信息干扰以及各种不确定性信息会对运行人员的分析判断造成巨大的困难。因此研究变电站的故障诊断模型及系统,为运行人员快速诊断出故障,并对故障恢复提供依据,对于保证整个联网工程的安全有效运行具有着重要的意义。论文主要研究内容如下:(1)本文针对已有Petri网在750kV变电站故障诊断研究中模型建立庞大、建立逻辑复杂、通用性不强等问题,并结合750k V变电站双网双保护的配置,提出了基于有色Petri网的750kV变电站故障诊断方法,该方法将同类型库所横向折叠和同结构变迁纵向压缩,同时将保护和断路器用单色及复合色进行颜色表示,从而使元件保护断路器信息分类处理,条理清晰。在模型和推理中采用了分层和分网的思想,将其划分为信息转化层、信息比较层、信息诊断层和信息整合层四层,使得模型建立和推理简单清晰,同时将其应用在主网和冗余网中,最后将二者耦合得到耦合网。通过仿真验证表明:该方法对750kV变电站单重和多重故障均能得到正确的结果,且推理速度快,模型简便,且其结构不受保护配置和电压等级的改变而改变,通用性较强。(2)针对有色Petri网在750kV变电站故障诊断中得不到完备信号时易出现诊断错误以及保护拒动误动时诊断的可信度下降无法体现出来,提出了一种基于改进有色Petri网的故障诊断方法,侧重于通过保护信息的分类并与断路器内在逻辑关系的转化得到真正表示断路器状态的信息,再与能够直接表征故障的保护信息综合诊断;同时在有色Petri网中引入模糊因子λ,相应地颜色属性中包含着置信度属性w,解决诊断可信度下降及征兆信息不确定性问题。通过变迁标识状态的演变进行分类将整个改进有色Petri网分层表示,逐层精确,求解主网和冗余网的诊断结果,并通过耦合得到最终诊断结果。通过对单重和多重故障的仿真,该方法故障征兆信息表述明确,模型结构简单、规模小、适用性强,并能较好的解决信息缺失、存在拒动误动等不确定性问题的处理。
[Abstract]:The Northwest 750kV interconnection project is an ultra-high voltage transmission and transformation project. It has the characteristics of large scale, long transmission line, high voltage, large units and close interconnection. It is the key part of transmitting various forms of electric energy in Northwest China. 750kV substation acts as its core link. The degree of intelligence is becoming higher and higher, and more monitoring devices are used. The topology and protection configuration are complex and cumbersome compared with the general substation and transmission network. Once failure occurs, there will be a large number of alarm information, and there are many kinds of information, redundant information interference and various uncertain information will cause great difficulties for operators to analyze and judge. Therefore, it is of great significance to study the fault diagnosis model and system of substation in order to quickly diagnose the fault and provide the basis for fault recovery, which is of great significance to ensure the safe and effective operation of the whole network project. The main contents of this paper are as follows: (1) aiming at the problems of establishing large model, complicated logic and weak generality in the fault diagnosis research of 750kV substation based on Petri net, combined with the configuration of double network and double protection in 750kV substation. A fault diagnosis method for 750kV substation based on colored Petri net is proposed. The method compresses the same type of library horizontally and vertically with the same structure transition, and represents the protection and circuit breaker with monochromatic and composite colors at the same time. Thus, the information of component protection circuit breaker can be classified and processed clearly. The idea of layer and network is adopted in the model and reasoning, which is divided into four layers: information transformation layer, information comparison layer, information diagnosis layer and information integration layer, which makes the model establishment and reasoning simple and clear. At the same time, it is applied to the main network and redundant network. Finally, the coupling network is obtained by coupling the two networks. The simulation results show that the method can get correct results for both single and multiple faults of 750 kV substation, and the reasoning speed is fast, the model is simple, and its structure is not changed by protection configuration and voltage grade. (2) in view of the fault diagnosis of 750kV substation, the colored Petri net is easy to appear the diagnosis error when the fault diagnosis is not complete, and the reliability of the diagnosis when the protection is not in the wrong operation can not be reflected. In this paper, a fault diagnosis method based on improved colored Petri nets is proposed, which focuses on the classification of protection information and the transformation of the inherent logic relationship with the circuit breaker to obtain the true information representing the state of the circuit breaker. At the same time, fuzzy factor 位 is introduced into colored Petri net, which contains confidence attribute w, which solves the problem of reliability decline and uncertainty of symptom information. By the evolution of transition identification states, the whole improved colored Petri net is represented by layers, and the diagnosis results of the main network and redundant net are solved, and the final diagnosis results are obtained by coupling. Through the simulation of single and multiple faults, the fault symptom information is expressed clearly, the model structure is simple, the scale is small, the applicability is strong, and it can solve the uncertain problems such as missing information and maloperation.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM63

【参考文献】